1. Field of the Invention
The present invention relates to a method for predicting the result of a race, in particular the arrival order of competitors participating in a race.
2. Description of the Related Art
Currently many races of human being and animals, such as a bicycle race, a boat race, a horse race, a dog race, etc., are held. Several prediction devices for predicting the results of these races are also provided. For the examples, there are prediction devices disclosed in Japanese Patent Laid-open Nos. 10-216355, 11-290553 and 11-290554. Patent Laid-open No. 10-216355 discloses predicting the arrival order of a race based on the capability value of racehorses. Both Patent Laid-open Nos. 11-290553 and 11-290554 disclose predicting the result of a horse race based on information about racehorses, such as a training result, running ability, pedigree, etc.
FIG. 1 shows the concept of the conventional prediction device. As shown in FIG. 1, the conventional prediction device predicts race results based on the characteristic peculiar to an individual racehorse (hereinafter called a xe2x80x9chorse characteristicxe2x80x9d) led by the capability, pedigree, running ability of an individual racehorse. Since this concept in the case of a racehorse also applies to other races, a factor related to a competitor, such as a racehorse, that is, running human being, animal, vehicle, etc., is called a competitor-related factor.
FIG. 2 is a flowchart showing the process of the conventional prediction device. The process of the conventional prediction device is described with reference to FIG. 2. First, the prediction device receives the designation of a race the result of which a user wants to obtain (hereinafter called a xe2x80x9ctarget racexe2x80x9d) from the user (step ST10). Then, the prediction device judges whether each competitor participating in a target race conforms to the competitor-related factor (step ST12). It is judged whether each racehorse is good at a target horse race ground, for example, good at a long lawn course or whether the father horse of each racehorse has won a major race. Information about the physical condition before a race of each racehorse is sometimes taken into consideration.
Based on the judgment result in step ST12, the prediction device attaches a score to each competitor (step ST14) and presents the result with a score to the user (step ST16). The user judges the arrival order of each competitor based on the result with a score. FIG. 3 shows an example of the conventional prediction result of a horse race. As shown in FIG. 3, a prediction result varies depending on a predictor. This is because a race result prediction is influenced by the predictor""s handling of each factor constituting a competitor-related factor or the predictor""s subjective importance degree of each factor.
As described above, the conventional prediction device has a problem that the prediction result of a race is influenced by a predictor""s subjectivity, that is, the handling of each factor constituting this competitor-related factor of the manufacturer and user of a prediction device. For example, in the case of a horse race, although a horse characteristic is composed of pedigree, a training result (physical condition), etc., the handling of pedigree and a physical condition, the subjective importance degree of these factors vary depending on the subjectivity of the manufacturer and user of a prediction device. That is, the prediction result of a race varies depending on the subjectivity of the manufacturer and user of a prediction device, which is a problem. This leads to the dispersion of prediction reliability.
In the conventional prediction device uses a competitor-related factor, but does not use a factor that is found when past race results are statistically processed, etc., which is another problem.
It is an object of the present invention to enable the use of the statistical result of past race results, which are rather objective, in addition to the analysis result of a competitor-related factor, which is influenced by the subjectivity of the manufacturer and user of a prediction device in view of the problems described above and eventually to improve prediction reliability.
As described above, the present invention is useful for the result prediction of a race, in particular when the arrival order of competitors participating a race is predicted.
According to the first aspect, the prediction device for predicting a race result comprises a statistical unit for statistically processing past race results with a condition related to the race condition of a target race and a prediction unit for predicting the result of a target race based on both the analysis result that is based on the characteristic of each competitor participating in the target race and the statistical result obtained by the statistical unit.
In this way, a race result can be predicted using the statistical result of past races, which is rather objective, in addition to the analysis result based on a competitor-related factor being the individual characteristic of each competitor, which has a disadvantage of being influenced by the subjectivity of the manufacturer, etc., of a prediction device. Eventually, the dispersion ratio of prediction reliability can be reduced and prediction reliability can be improved.
In the configuration the statistical unit can also comprise a race condition extraction unit for extracting the past race results with the same race condition as that of the target race. In that case, by extracting a past race result with the same race condition as that of a target race and taking statistics of these extracted past race results, statistical data can be made appropriate and as a result, a more reliable statistical result can be obtained. In this case, for a race condition, information about the place, time and category of a race can also be used.
Also, in the configuration, the statistical unit can comprise a factor extraction unit for extracting an effective factor, which is related to arrival order, from past race results by sorting the past race results according to arrival order and a factor conformation judgment unit for judging whether each competitor participating in a target race conforms to the extracted effective factor and attaching information about the judgment result to each competitor.
By sorting past race results according to arrival order, an effective factor related to arrival order can be obtained. This effective factor can be expected to be useful for the prediction of arrival order. For example, in the case of a horse race it is assumed that most of the racehorses that have won in past horse races have a tendency to xe2x80x9close three or more kilograms of weight before the racexe2x80x9d. Then, it can be expected that the xe2x80x9closs of three or more of weightxe2x80x9d is an effective factor related to arrival order and can be useful for the prediction of arrival order. In this case, it is judged whether each racehorse participating in a target race conforms to the effective factor of xe2x80x9closing three or more kilograms in weightxe2x80x9d, and a score is attached to a satisfied competitor. In this way, a racehorse with the same tendency as that of a racehorse that has won in past horse races can be found out of racehorses participating in a target race. By repeating such judgment for several effective factors, a racehorse having the high possibility of wining the race can be statistically found.
An effective factor can be obtained by sorting the items of past race results stored in a competition tendency-related factor storage unit and extracting an item with a prescribed tendency which most of competitors that have obtained good results show. Such an effective factor can be used to find a competitor having a high possibility of obtaining a good result.
Conversely, by extracting an item with a prescribed tendency which most of competitors that have not obtained good results show, an effective factor can also be obtained. Such an effective factor can be used to find a competitor having a low possibility of obtaining a good result.
Further, in the configuration, the prediction unit can predict the result of a target race based on both the analysis result that is based on the characteristic of each competitor participating in the target race and the importance degree, which indicates the degree of importance (weight) of the statistical result of past race results. In this case, by setting the importance degree, a user can determine the importance degree of both the analysis result on each competitor and the statistical result of past race results.
In the configuration, a prediction device can also be a network terminal and can receive past race results to be used for statistics via a network.
Another aspect of the present invention is a prediction method for predicting a race result. In this case, past race results with a condition related to the race condition of a target race are statistically processed, and the result of the target race is predicted based on both the analysis result that is based on the characteristic of each competitor participating in the target race and the statistical result. In this way too, the problems described above can be solved.
The problems can also be solved by making a computer to read a program for enabling a computer to exercise the same control as the function performed by each configuration described above, from a computer-readable storage medium on which is recorded the program and to execute the program.